One simple New Keynesian model would be http://crei.cat/people/gali/pdf_files/monograph/slides-ch3.pdf but I do not restrict my question to this model only.

Let us for now ignore the problem of linearizing around zero-growth-rate, and assume that there is no problem in modelling constant natural growth rate.

Let us suppose that monetary shock occurred, due to stochastic or forecast errors or intentional deviation. Can this shock result in technology shock also? (ex. change in $A_t$ in production function?)

I am sure that there exists a model that does this, but in most models, and especially Gali's model as linked above, is this true?

  • $\begingroup$ is there a reason why you keep making new accounts? $\endgroup$ Dec 18, 2014 at 1:36

1 Answer 1


No it usually does not.

There is an intuitive reason why: We interpret the technology shock as a residual, the unexplainable part in TFP changes. If we knew that there was a specific causal relationship between monetary policy and future TFP, that part of future TFP would now was explained.

It's true that the MP shock is excluded in standard Solow Regressions, but if you believe there is a causal effect, you need to regress it out of it. Since after all, we understand $A_t$ to be independent.

Note that still, implicitly, through its effects on capital accumulation (which does not exist in standard NK models), monetary policy should impact future TFP. Then, you could make $A_t$ dependent and argue it being a shortcut in the absence of capital.

  • $\begingroup$ In most standard models, capital accumulation doesn't impact TFP either. I guess what you mean is that more capital makes average product of labor higher, and in a model without capital, the A_t is also average labor productivity, so having it depend on monetary shock might capture some effects of unmodeled response of capital. But I think it would be simpler to just add capital to the model. $\endgroup$
    – ivansml
    Dec 17, 2014 at 16:41
  • $\begingroup$ Right. But adding capital in a NK world is actually quite complicated. $\endgroup$
    – FooBar
    Dec 17, 2014 at 17:00
  • $\begingroup$ Depends on how it is modelled. Larger-scale models like Smets-Wouters routinely do include capital, which is owned and accumulated by households and rent to intermediate-good firms each period. But if each firm owned its own capital stock and couldn't reallocate easily to other firms, then yes, that would require modelling the heterogeneity explicitly and make the model much more complicated. $\endgroup$
    – ivansml
    Dec 17, 2014 at 18:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.